Leading Russian bank develops facial-recognition network for finding lost pets, with 94% accuracy for dogs & 84% for cats
Speaking to news agency TASS, Sberbank’s press service explained that artificial intelligence makes it possible to spot differences between individuals of the same species.
"Sber has taught a neural network to distinguish very similar animals by certain features of their faces,” the press service said. “At the moment, the quality of comparison of two pet faces is 94% for dogs and 84% for cats.”Also on rt.com Russia registers world’s first Covid-19 vaccine for cats & dogs as makers of Sputnik V warn pets & farm animals could spread virus
The system is based on the CosFace algorithm, which is used to recognize human faces.
In 2014, Sberbank made international headlines when it began lending cats to clients who got a mortgage through the company. According to Russian tradition, feline friends are a sign of good luck when moving into a new home.
If you like this story, share it with a friend!